The Research On Fault Diagnose For Truck Scale Multi-sensor System

Truck scale is widely applied in many fields such as storage, trade, transport, communication, industry and mine. Weighing sensor is one of the core components in a truck scale, which is in charge of converting load weight into electrical signals. In addition, the working environment is too bad and the number is more, these reasons lead to high probability of weighing sensor’s fault. At the same time, virtual instrument technology has become the developing tendency of measurement field with the fast development of computer hardware and software. In this paper, the use of the advantage of virtual instrument and the combination of theories and methods of fault diagnosis were made to develop the system of truck scale’s multiple load cells fault detecting and diagnosis.Firstly, the research status and development trends of truck scale are discussed. The concept of fault diagnosis was proposed. The problem in existing methods on the sensors’fault detective and diagnosis are summarized.Secondly, the composition of truck scale and their principle are introduced. The weighing principle of truck scale is expounded. The principle of load cells in parallel mode is analyzed and its characteristic is compared. The composition and technical characteristics of intelligent truck scale is introduced briefly.Thirdly, the BP Neural Networks fault diagnosis for truck scale’s multi-sensor weighing system is studied. For truck scale’s fault-tolerance and load cell’s fault diagnosis, the failure cause and failure type of load cell is analyzed. To the characteristic of multi-sensor system, using the nonlinear approach ability of BP Neural Networks, makes the twice prediction outputs of the sensor that will be diagnosed. The first time prediction is used to identify the failure; the second time prediction is used to locate the fault sensor and makes use of the first time predictive output data to resume the fault signal.Fourthly, multi-sensor fault diagnosis system base on LabVIEW platform is studied. The system realized offline-training for the BP Neural Network based time series predictor and multi-sensor’s online detecting. Moreover, through the system, optimum parameters of models can be saved; signal of Weighing Sensor can be displayed; fault alarm function is realized.Finally, the performances of intelligent truck scale with the target fault sensor, such as eccentric error, linearity, repeatability, discrimination and zero error, are verified in field. Field test shows multi-sensor fault diagnosis system for intelligent truck scale runs reliably.